Collecting Typhoon Disaster Information from Twitter Based on Query Expansion

Zi Chen,Samsung Lim

doi:10.3390/ijgi7040139

Abstract

Social media is a popular source of volunteered geographic information owing to its massive real-time data; however, the use of social media data in the context of geospatial analysis is challenging because complex semantic filters are required for the aggregation of geographic messages from the data streams. This article proposes a new query expansion method for social media streams which updates the query keywords periodically by the words extracted from the preceding search results. The proposed method has optimized the trade-off between precision and coverage of geographical messages by factoring in the influences of the keyword number and refresh cycle in the query process, and some improvements on the classic Term Frequency-Inverse Document Frequency (TF-IDF) method for short texts were achieved. Furthermore, a number of filters based upon relevance to the target topic were established and tested. This method was tested on a dataset from Twitter within the geographic extent of Macau in August 2017 during two consecutive typhoon hits. The result supports its effectiveness with a controllable precision and considerable increment of relevant information. Moreover, the query keywords can adjust themselves to the local language environment by discovering new keywords. To conclude, this query expansion method is able to provide a reliable method for social media-based information retrieval.

Highlights

Volunteered geographic information (VGI) [1] has expanded the sources of geographic information from experts to general public, and shifted geographic information systems (GIS) from an abstruse technology to a medium of communication [2]
The total message number (TMN) represents the amount of data to be searched in the time span of the current round, and the search results are the quantities of positive messages
In Rounds 2 and 3, the numbers of search results soared along with the increase in the TMN, and many symbolic keywords related to typhoons occur, such as ‘signals’

Summary

Introduction

Volunteered geographic information (VGI) [1] has expanded the sources of geographic information from experts to general public, and shifted geographic information systems (GIS) from an abstruse technology to a medium of communication [2]. The process of social media data collection consists in searching by a couple of fixed keywords or hashtags to filter relevant information [7,8,9,10] These keywords or hashtags, which normally define or summarize the topic or incident, are usually manually selected to ensure a high precision of search results and remain unchanged throughout the search process. This intuitive approach may cause omission of a massive amount of relevant information if the quantity of the keywords is limited or the keywords are not appropriately chosen, which leads to incomplete information retrieval. As the discussion of the topic develops on social media, the discussion focus may deviate from the initial keyword set, or the keywords become too general to cover a varying corpus of the topic, in a different language environment; as a result, a high volume of irrelevant information can be collected

Objectives

Methods

Results

Conclusion